Smart Home for Elderly: Modeling and Simulationceur-ws.org/Vol-1294/paper17.pdf · Smart Home for Elderly: Modeling and Simulation . Yacine Kissoum . Faculty of Sciences, Computer

Smart Home for Elderly:

Modeling and Simulation

Yacine Kissoum

Faculty of Sciences, Computer Science

Department

Université 20 Août 1955, Skikda.

Algeria

[email protected]

Sara Kerraoui


Department


Algeria

[email protected]

Mohamed Lamine Boughaouas


Department


Algeria

[email protected]

Abstract – The development trend of computing is that the intelligence will become ambient. It is a one-user multi-device paradigm. These devices are transparent to the user and they can react in a proactive and non-intrusive manner. Moreover, among the general population those most likely to benefit from the development of this technology are the elderly and dependent people. They often wish to continue living independently in their home as opposed to being forced to live in a hospital. For that, homes should be smart in order to provide safety, security to reduce falls, disability, stress, fear or social isolation. Nevertheless, much of processes used for smart home design are ad-hoc. To cope with this drawback, while taking into account both dimensions of time and space, it is required to use a suitable formal model that is able to handle smart home domain specific nature. This paper proposes a technique that uses reference nets to model and simulate a smart home which is sensitive, adaptive and responsive to elderly’s health, needs and preferences.

Keywords – Ambient intelligence, Multi agent systems, Nets within nets, Elderly, Timed nets, Smart home.

1. INTRODUCTION

In recent years, there has been an important growth in

the field of Ambient Intelligence (AmI) [5, 6],

involving major changes in the daily lives of people.

The vision of Ambient Intelligence implies a

seamless environment of computing and advanced

networking technology that is aware of human

presence, personalities, needs and which is capable of

responding intelligently to spoken or gestured

indications of desire, and even in engaging in

intelligent dialogue. However, the development of

systems that clearly fulfill the needs of AmI is

difficult and not always satisfactory. It requires a

joint development of methods, techniques and

technologies based on services. An AmI based

system consists on a set of human actors and adaptive

mechanisms which work together in a distributed

way. Those mechanisms provide on demand

personalized services and stimulate users through

their environment according specific situation

characteristics [1].

Moreover, the percentage of elderly in today’s

societies keeps on growing. With the current trends in

population demographics, it is becoming increasingly

difficult for governments worldwide to fully support

the health and social care systems [11]. The use of

smart technologies, including smart homes could

arguably relieve the pressure on aged care health and

social support services [10]. The challenge with

smart home technologies for elderly is to create a

home environment that is safe and secure to reduce

falls, disability, stress, fear or social isolation [4].

Like any other system, smart home development

starts with a high level model and proceeds through a

process of refinement, simulation, verification,

implementation and test. Much of processes used for

ICAASE'2014 Smart Home for Elderly: Modeling and Simulation

International Conference on Advanced Aspects of Software EngineeringICAASE, November, 2-4, 2014, Constantine, Algeria. 148

smart home design are ad-hoc. By time, all the gritty

details of implementation are taken care with the

original system description has pretty much been lost,

causing a lack of design oversight and a surplus of

one-time-only design artifacts [2]. Besides, very little

can be done within a smart home system without an

explicit or implicit reference to where and when the

meaningful events occurred. For a system to make

sensible decisions it has to be aware of where the

inhabitants are and have been during some period of

time. These insights, together with other information,

will provide important clues on the type of activities

the user is engaged in and the most adequate

response.

So, to cope with the ad-hoc nature of smart home

systems design process and to take into account both

dimensions, space and time, to understand key

elements of a situation under development, it’s

required to use a suitable formal model that is able to

handle smart home domain specific nature. Thanks to

Köhler et al [6] who proposed a model that deals with

complex systems in an elegant and intuitive manner

without losing formal accuracy. A model built upon a

formalism that has a formal semantics to support

verification, timed simulation and execution. The

proposed approach for modeling smart home

proposed so far is based on such a model called

Reference net [10]. Reference nets are based on the

nets within nets paradigm that generalizes token to

data types and even nets.

The paper is structured as follows: section 2 surveys

related work. In section 3 we give a short

introduction on the paradigm of reference net and

show how this paradigm can be used to model the

mobility concept in multi-agent special case. Section

4 describes the case study which will be modeled in

section 5; Sections 6 discusses the result of

simulation. Finally the section 7 concludes this work.

2. RELATED WORK

Smart home are often the ideal solution for

individuals with different needs and abilities. This is

why several projects were introduced in recent years.

Pan et al. [5] for example presented a homecare

service that tackles the problem of fall discovery

using a tri-axial accelerometer and reporting such a

discovery to an emergency center. The paper

introduces a fall detector based on a neural network

and a multi-agent architecture for requesting

emergency services. The Aging in place project1 from

the University of Missouri gives seniors an

1 http : //aginginplace.missouri.edu/

independent space that allows guaranteeing adequate

medical supervision with dedicated sensors. Soprano-

UE project2 (Services-Oriented programmable Smart

environments for older Europeans), funded by the

European commission, aims to help older people to

have a more independent life. The main contribution

of the project was the creation of middleware

OpenAAL for software implementation of smart

homes.This is an open software framework that takes

as input the information from sensors installed in the

house and that the active services to perform actions

on the environment. The originality of this system is

to provide two layers for semantic representations of

the environment: the first is a description of sensors

and their states, and the second describes the situation

of the inhabitants, activities, location, and

emergency. Another project is the DAT project [3]

that aims at building up a smart home environment

where people with disabilities can improve their

abilities to cope with daily life activities by means of

technologically advanced home automation solutions.

The project has a threefold purpose. The smart home

will be used as a physical setting, where clients with

disabilities can follow individual programs aimed at

improving their independence in the home

environment.

3. TIMED NETS AND MOBILITY

Reference nets [12] are a graphical notation that are

especially well suited for the description and

execution of complex, concurrent processes. As for

other net formalisms, there exist tools for the

simulation of reference nets called Renew (for

Reference Net workshop). Reference nets extend

black and colored Petri nets by means of net

instances, nets as token objects, communication via

synchronous channels, and different arc types.

Definitions of these extensions are given in [8, 9].

While pure Petri nets capture the causality and

conflict situations of a system nicely, there are

reasons to add a notion of time to the formalism in

order to model additional dependencies. This is

especially true in the case of simulations of real-

world systems.

In timed nets, a time stamp is attached to each token.

It denotes the time when the token becomes

available. Delays may be used with arcs in order to

control the time stamps of token and the firing times

of transitions. A delay is added to an arc by adding to

the arc inscription the symbol @ and an expression

that evaluates to the number of time units. For

2 http://www.soprano-ip.org/


International Conference on Advanced Aspects of Software Engineering ICAASE, November, 2-4, 2014, Constantine, Algeria. 149

example, x@t indicates that the token value x has to

be move after t time units.

Look to Figure 1. This is an example model of a

simple traffic light where yellow light is omitted. On

the left hand side net, the token in the place green has

to move to the yellow place after 30 units of time.

This will synchronize the transition inscribed with by

the channel this:go1() which make the token on the

place red (on the right hand side net) moving to the

green place. Subsequently after 30 units of time, the

transition inscribed by the channel this:go2() on the

right hand side net become enabled and the traffic

light will switch to red. Note that cars, modeled as

mobile agents, are created every 10 units of time on

the lower nets (a:new cars_1() and b:new car_2()),

each being numbered accordingly. Cars on either side

cannot move only if the transition isL1Green

(respectively isL2Green) is enabled. Such transitions

are modeled using test arcs. Test arcs, which have no

arrowheads, allow a single token to be accessed

(tested) by several test arcs at once.

Figure 1: The traffic light example.

This example gives an idea how the interplay

between object net (token net) and system net can be

used to model mobile entities moving through a

system net. Such a net offers or denies possibilities to

move around, while the mobile object net moves at

the right time by activating respectively, the

transition that is inscribed with the precondition of

the channel, and the move transition in the system

net. Without the viewpoint of nets as tokens, the

modeler would have to encode the mobile agent as a

data structure for example. As a consequence, the

inner actions of the mobile entity cannot be modeled

directly, so they have to be lifted up to the system

net, which seems quite unnatural. By using nets

within nets we can investigate the concurrency of the

system and the mobile agent in one model without

losing the needed abstraction [10].

To investigate this modeling method, below we

introduce the smart home for elderly case study. Then

we show how we model this example by means of

reference net paradigm.

4. SMART HOME FOR ELDERLY CASE

STUDY

Let us imagine a house with several rooms: living

room, kitchen, bed room, next room, bath room and

the front yard. To be smart, such a house should

contain a large set of services that cooperate to

simplify the life of the owner, to make energy saving,

and to provide comfort and security solutions. Figure

2 show an example of a smart home where various

home devices are scattered over rooms to constitute a

cooperating environment that provide smart services

to the owner.

Figure 2: The smart home case study.

More precisely, to ensure security and safety of the

owner the smart home uses information from

temperature level sensors (Tl), temperature rate

sensors (Tr) and smoke detectors (S) to detect

possible fire and to take adequate actions. In the

kitchen, where there is a possibility of gas leakage, a

gas detector is installed (G). Gas leakage keeps an

open eye to detect any gas leakage and take

immediate action. There are image sensors (I) and

motion detectors (M) in all places to identify any

visitor and check whether he is authorized or no.

Glass break sensors (B) and contact switch (C) are

also installed on all windows/doors to monitor their

safety. In addition, smart home provides energy

saving solutions through installing occupancy sensors

(O) in all rooms. Systems like lighting (L) will

consider environmental conditions for their operation

to help saving more energy.

The overall system will be designed according to

MULAN [7] architecture. MULAN is implemented

in RENEW3 and has the general structure as depicted

in Figure 3 Each box describes one level of

abstraction in terms of a system net.

3 www.renew.de



The net in the upper left side of Figure 3 describes an

agent system, which places contain agent platforms

as tokens. The transitions describe communication or

mobility channels, which build up the infrastructure.

By zooming into the platform token on place p3, the

structure of a platform becomes visible. The central

place agents host all agents, which are currently on

this platform. Each platform offers services to the

agents. Agents can be created (transition new) or

destroyed (transition destroy). Agents can

communicate by message exchange. Two agents of

the same platform can communicate by the transition

internal communication. External communication

only binds one agent, since the other agent is bound

on a second platform somewhere else in the agent

system. Also mobility facilities are provided on a

platform: agents can leave the platform via the

transition send agent or enter the platform via the

transition receive agent.

Agents are also modeled in terms of nets. They are

encapsulated, since the only way of interaction is by

message passing. Agents can be intelligent, since

they have access to a knowledge base. The behavior

of the agent is described in terms of protocols, which

are again nets. Protocols are located as templates on

the place protocols. Protocol templates can be

instantiated, which happens for example if a message

arrives. An instantiated protocol is part of a

conversation and lies in the place conversations.

Figure 3: Agent systems as nets within nets (adapted from

[9])

5. MODELING SMART HOME COMPONENTS

Using nets within nets as a modeling paradigm

allows for the direct use of system models at

execution time. This can be exploited as follows: The

overall system will be designed as a system net with

places defining locations (rooms) in or in front of the

house: Transitions model possible movements

between these rooms (green transitions), Figure 4.

Due to the paper size limitation, only a subset of

services will be presented here to highlight the

modeling concept. The idea can be further extended

to any smart home room and to any additional

services.

Figure 4: The smart home system net.

To be smart, rooms should be sensible to the human

presence. Being sensible demands recognizing the

user, learning or knowing her/his preferences, and the

capability to exhibit empathy with or react to the

user’s mood and the prevailing situation. This is why

we have attached to each room a cognitive stationary

agent (blue transitions). These agents are: the hall

agent, the kitchen agent, the living room agent, the

bed room agents, the next room agent and the bath

room agent. Also, from all the electronic devices

scattered over the house we have selected those

laying in the living room and in the kitchen, namely:

the TV agent, the AC agent, the cooker agent and gas

leakage detector which are modeled again as agents

(brown transitions). There is also one particular agent

(red transition) responsible of checking the blood

pressure and/or glucose rate of the elderly at random

times of a day. It may take the form of a bracelet

hanging on the wrist of the elderly. Finally, this

model of house is filled with life by implementing

human house inhabitants (yellow transitions). These

ones are the owner (the elderly); his doctor and let us

say his son (the creation and arc inscriptions of these

two agents will be detailed so far). All these agents

share the same structure depicted in lower right side

of Figure 3 (of course their knowledge bases and

protocols are different).

The behaviors of these agents are described in terms

of protocols. The protocol selection can basically be

performed pro-actively or reactively. This distinction

corresponds to the bilateral access to the place

holding the protocols: protocols. The only difference

in enabling and occurrence of the transitions rea and

pro is the arc from the place incoming messages to



the transition rea (see agent box of Figure 3). So it

may only be enabled by incoming messages. Both the

reaction to arriving messages and the commencement

of a new conversation is influenced by the knowledge

base place.

Human preferences and moods are part of these

agent’s knowledge bases. In simple cases the

knowledge base place can be implemented for

example as subnets. This is true in the case of sensors

(modeled as reactive agents). Unfortunately, using

such implementation for cognitive agents leads to a

closed model that is often needed to rethink

completely before introducing any modification. For

this kind of agent advanced implementations of the

knowledge place as the connections to an inference

engine are also possible. In this way, we will make

use of Java Expert System Shell4 (JESS). As for other

expert systems, JESS is composed from a rule base

which corresponds to the knowledge base and a

working memory which corresponds to the fact base.

Its inference engine is composed from the pattern

matcher which decides what rules to fire and when

and the agenda which schedules the order in which

activated rules will fire. The execution engine is

responsible for firing rules and executing other code.

At home, the behavior of the owner is highly

dependent on the timeframe in which he is located.

Intuitively these slots are morning, noon, afternoon

and evening. Thus, different scenarios related to such

parts of the day can be imagined. The ‘’wakeup

mode’’, the ‘’leave home mode’’, the ‘’return home

mode’’, the ‘’evening mode‘’ and the ‘’go to sleep

mode’’ are examples of these scenarios.

Before discussing one of these scenarios, we start by

defining the working memory and the rule bases of

each our agents. A particular attention will be given

to the living room and the kitchen agents. Indeed,

such agents must keep an open eye on the elderly

living in this house. For this, it is necessary that they

store information about this person, its environment

as well as other agents. Also, they must exchange

their respective information to ensure the welfare of

the elderly. For that, each JESS rule engine holds a

collection of knowledge nuggets called facts. Every

fact has a template. The template has a name and a

set of slots, and each fact gets these things from its

template. We start by defining the ACL message

template as following:

(deftemplate ACLMessage (slot

communicative-act) (slot sender) (slot

receiver) (slot conversation-id) (slot

protocol) (slot language) (slot

ontology) (slot content))

4 http://herzberg.ca.sandia.gov

It is important to note that elderly preferences

described in each agent’s working memory, depend

on the room in which he is. It will be favorite

program TV or multimedia playlist for the living

room agent, water temperature for the bath room

agent, light dimming and air conditioner levels for

the bed room agent. The following templates define

the owner of the house (the elderly) and people close

to him.

(deftemplate elderly (slot name) (slot

sex) (slot age) (multislot disease)

(multislot TVprefer) (multislot

Musicprefer))

(deftemplate related (slot name) (slot

sex) (slot age) (slot type) (slot

telephone))

Other facts necessary to our room’s agents are given

in the following. They let them, among others, to

create the desired ambiance once the owner enters the

living room or the kitchen.

(deftemplate enter (slot name) (slot

room) (slot time))

(deftemplate television (slot state)

(slot channel))

(deftemplate airCond (slot state) (slot

temperature))

(deftemplate coffeeMaker (slot drink)

(slot sugar))

Also, and because this house is owned by an elderly

suffering from one or more diseases (usually

chronic), agents should be able to analyze the

significant factors related to these diseases.

Therefore, our agents will use the following

templates:

(deftemplate bloodpressure (slot

maxVal) (slot minVal))

(deftemplate glucose (slot rate) (slot

state))

Finally, we must note that in a smart home, devices

are often used to build an environment in which

many features of the home are automated. But in

some situations the user decides whether to perform

tasks alone (manual mode) or allow devices

throughout the home to realize its tasks (automatic

mode).

Assigning values to these slots allows room’s agents

to identify persons having access to the house. An

example of authorized persons is as follows:

(assert (elderly (name Ben) (sex male)

(age 69) (disease Diabetes

Hypertension) (TVprefer News Docs

Entertainment) (foodprefer tea fish

salad))

(assert (related (name Tim) (sex male)

(age 45) (type doctor) (telephone

077777777))



(assert (related (name Bob) (sex male)

(age 35) (type son) (telephone

06666666))

Now we can define some JESS rules. Every JESS

rule has two parts, separated by the "=>" symbol. The

first part consists of the LHS pattern. The second part

consists of the RHS action. The LHS of a rule

consists of patterns which are used to match facts in

the working memory, while the RHS contains

function calls. Examples of such rules are the

following. In the first one, as soon as the owner

enters the living room, the agent will check the

current time (morning in this case) then turn on the

television and air conditioner with the owner’s

desired channel and temperature. In the second, the

kitchen agent will ask the coffee machine to prepare a

tea without sugar.

1 :(defrule rule1

(enter {name == Ben && room == living

room && time >= 8 && time <= 10})

(elderly {Tvpref == news})

=>

(assert (television (state ON) (channel

NSN)))

(assert (airCond (state ON)

(temperature 25)))

2 :(defrule rule2

(enter {name == Ben && room == kitchen

&& time > 10 && time <= 12})

(elderly {foodpref == Tea})

=>

(assert (coffeeMaker (drink Tea) (sugar

No)))

In addition, the following rules allow

the bracelet agent to decide on the

health of the elderly based on his

blood glucose.

3 :(defrule rule3

(glucose {rate < 0.60 })

=>

(assert (ACLMessage (communicative-act

INFORM) (sender wrAg) (receiver lrAg)

(content "hypoglycemia ")

(conversation-id "call son"))

4 :(defrule rule4

(glucose {rate > 0.70 && rate <1.10 &&

state == fasting })

=>


INFORM) (sender wrAg) (receiver kiAg)

(content " normoglycemia "))

5 :(defrule rule6

(glucose {rate > 1.10 })

=>


INFORM) (sender wrAg) (receiver kiAg)

(content " hyperglycemia ")

(conversation-id "call doctor"))

Let us focus on the elderly agent. Suppose that we are

in the “wakeup mode”, the execution of the of the

JESS program laying in the elderly agent knowledge

place (modeled in our case as mobile agent)

concludes with a valid fact “take breakfast”. Note

that both JESS and Renew are written in Java. More

precisely, references net are themselves Java objects.

Making calls from JESS code to nets is just as easy as

to make calls from nets to JESS code. So, Driven by

this fact, the elderly agent selects a plan to execute:

“enter kitchen”. Actually the plan execution

corresponds to the selection and the instantiation of

the exact protocol and commencement of the

conversation. Figure 5 shows such a protocol.

Figure 5: The enter kitchen protocol.

This protocol is selected pro-actively. After its

instantiation, the transition mv(“kt”) produces a

performative i that define the entering agent’s

identity, and which is directed to kitchen agent main

interface over the k:out(i) channel; subsequently the

agent terminates (by enabling the :stop() transition).

The k:out(i) is met by synchronizing with the same

transition in the kitchen agent net (:in(i)) which

induce the enabling of transition rea on agent main

net. Influenced by the identity of the person entering

the kitchen and the inference results provided by the

JESS program laying in knowledge base of the

kitchen agent, this one will instantiate, first, the

“welcome” protocol. So, knowing the entering

person’s identity and preferences, this protocol will

welcomes the entering person, then verify whether

the manual or the automatic mode is selected

(transition h:isManual()), in the first case, the kitchen

agent will do nothing but keep a close eye on the

elderly. In the second, it produces a performative d

defining the kind of the drink to prepare and over the

channel c:out(d), it will send it the ‘’request drink’’

protocol (Figure 6). With the same manner this

protocol terminates by enabling the transition :stop().

Figure 6: The welcome protocol.

After that this agent will instantiate the “request

drink” protocol (the automatic mode). So receiving

the kind of required drink, the transition compose

produces a performative containing a string that is



directed over the channel c:out(s) to the coffee maker

agent; subsequently the protocol is blocked waiting

for an answer message. An arriving answer enables

the transition answer. After occurrence of such a

transition the protocol is not blocked any further, and

the receiver will process the answer. Two situations

are possible: if the received answer is positive, then

the agent will update his beliefs; otherwise the

kitchen agent will undertake adequate actions. In

either case the protocol instance is deleted (by

enabling the stop transition), Figure 7.

Figure 7: The request drink protocol.

Figure 8 shows the prepare drink protocol. The

transition :in(dr) is responsible of passing the kind of

drink required (tea, coffee, or milk). Then the

protocol starts by verifying randomly if the required

quantity for the drink preparation is less to the

available one. If yes, the prepare transition is enabled

(note that drink preparation may take five minutes on

average). Otherwise, the transition supply is executed

and the transition :out(res) will inform the sender

(the kitchen agent) that his request is fulfilled or not.

Figure 8: The prepare drink protocol

The above protocols describe how agents collaborate

to assist elderly in the wake up scenario. Let us see

how these agents take care of elderly health.

Remember the bracelet agent that checks the glucose

rate of the elderly at random time of a day. It is

modeled as a mobile agent (having the same structure

as other agents) that is transported by the elderly

agent. Now, suppose that the manual mode is

activated, the cooker state is on and the bracelet agent

has detected a hyperglycemia. Once received, the

room agent in where the elderly is, instantiates the

protocol depicted in Figure 9. This one starts by

executing the transition inscribed by the channels

callSon and callDoctor. It is modeled to meet by

synchronization the same transitions in the house net

(see Figure 4). The doctor, for example, takes three

hours on average with a negative-exponential

distribution to reach the front yard place. After that,

the agent will check whether the manual mode is

activated or not. If so, the agent will switch to

automatic mode so it can broadcast safety messages

to other agents by enabling the transition this:Safety()

(turn off the cooker, the TV, the alarm, etc.).

Figure 9: The emergency protocol.

6. THE SIMULATION

Having defined the features of the smart home

components, the system is now ready for execution

(simulation). Figure 10 shows an ordinary simulation

state. Only the living room and the hall places are

depicted. In addition the living room agent net and its

protocol, the window in the background contains the

JESS facts and rules representing the knowledge base

of the living room agent.

It is now easy to see the state of the overall system by

looking at the system net. In fact, it is possible to take

control of the whole agent system implementing the

living room agent scenario by just double clicking on

one of the corresponding token. The tool RENEW

will display the respective net allowing for a

complete inspection of the running system without

having to interrupt it. The advantage is twofold [10].

Whereas a normal modeling process requires at

least three stages to reach an executable

program: (a) model the system, (b) implement

the model and (c) write the visualization for the

program, using the nets within nets paradigm,

the modeling process concludes with a running

system model.

The visualization of the system model at

execution time is indeed the implementation of

the system. This eliminates several potential

sources of errors shifting from model to

implementation to visualization in an ordinary

software design process.

7. CONCLUSION

In this paper we have presented an approach for

modeling of cooperating ambient agent cohabiting an

elderly’s home. An important argument for using the

reference nets paradigm is its strong expressiveness,

openness and versatility without losing formal

accuracy. Indeed, using such a paradigm, the

modeling process concludes with a running system



model. Also, this paradigm is selected not only for

mobility purposes in multi-agent systems, but mainly

because it is built upon a formalism that has a formal

semantics to support verification, execution and

timed simulation. We think that such robust and easy-

to-use tool reduces considerably the large initial

effort in term of man-hours required mainly in

constructing and validating smart home models.

Additional services such as fire system, curtain

control, climate control system, multimedia control

system and so on, will be directly integrated in the

design of the reference net based multi agent system

architecture MULAN.

8. REFERENCES

[1] A. Ricci, Buda, C. and Zaghini, N. An agent-

oriented programming model for SOA & web

services. In 5th IEEE International Conference on

industrial Informatics (INDIN'07),Vienna, Austria.

pp. 1059-1064, 2007

[2] A. K. Nabih, M. M. Gomaa, H. S. Osman and G.

M. Aly, Modeling, Simulation and control of smart

home Using Petri Nets, International Journal of

Smart home. Vol. 5 N° 3, July 2011

[3] A. Renzo, V. Gower, A. Caracciolo, G. Del

Zanna and M. Di Rienzo, The DAT Project: A Smart

Home Environment for People with Disabilities.

LNCS 4061, pp.492-499, springer –Verlag Berlin

Heidelberg 2006.

[4] J. Barlow, Venables T, Will technological

innovation create the true lifetime home? Housing

Studies 19: 795-810. 2004

[5] J.-I. Pan, Yung, C.-Y., Liang, C.-C., Lai, L.-F.:

An intelligent homecare emergency service system

for elder falling. In: World Congress on Medical

Physics and Biomedical Engineering, pp. 424--428.

Springer, 2006.

[6] M. Anastasopoulos, D., Bartelt, C., Koch, J. &

Rausch, A. Towards a Reference Middleware

Architecture for Ambient Intelligence Systems. ACM

Conference on Object-Oriented Programming,

Systems, Languages, and Applications. 2005

[7] M. Duvigneau, D. Moldt and H. Rölke,

Concurrent architecture for a multi-agent platform, in

Proceedings of the Workshop on Agent Oriented

Software Engineering (AOSE’02), LNCS 2585,

Springer Verlag, 2003.

[8] M. Köhler, D. Moldt, and H. Rölke, Modeling the

structure and behavior of petri net agents, J.M.

ICATPN 2001, LNCS 2075, 2001, 224–241.

[9] M. Köhler, D. Moldt and H. Rölke, Modelling

mobility and mobile agents using nets within nets,

ICATPN 2003, LNCS 2679, 2003, 121–139.

[10] M. Morris, Ozanne E, Miller K, Santamaria N,

Pearce A, Smart technologies for older people: A

systematic literature review of smart technologies

that promote health and wellbeing of older people

living at home. IBES, The University of Melbourne,

Australia. 2012.

[11] Agoulmine, Deen MJ, Jeong-Soo L, Meyyappan

M, U-health smart home: Innovative solutions for the

management of the elderly and chronic diseases.

IEEE Nanotechnology Magazine 5: 6-11. 2011

[12] O. Kummer, Simulating synchronous channels

and net instances. Workshop Algorithmen und

Werkzeuge für Petrinetze 1998, 73–78.

Figure 10: An example of simulation.



Documents

Smart Home for Elderly: Modeling and Simulationceur-ws.org/Vol-1294/paper17.pdf · Smart Home for Elderly: Modeling and Simulation . Yacine Kissoum . Faculty of Sciences, Computer